Package com.codedstreams.otterstreams.sql.runtime

Class TensorFlowSavedModelEngine

java.lang.Object

com.codedstream.otterstream.inference.engine.LocalInferenceEngine<org.tensorflow.SavedModelBundle>

com.codedstreams.otterstreams.sql.runtime.TensorFlowSavedModelEngine

All Implemented Interfaces:

InferenceEngine<org.tensorflow.SavedModelBundle>

public class TensorFlowSavedModelEngine
extends LocalInferenceEngine<org.tensorflow.SavedModelBundle>

TensorFlow SavedModel inference engine.

Nested Class Summary

Nested classes/interfaces inherited from interface com.codedstream.otterstream.inference.engine.InferenceEngine

InferenceEngine.EngineCapabilities

Field Summary

Fields inherited from class com.codedstream.otterstream.inference.engine.LocalInferenceEngine

initialized, loadedModel, modelConfig, modelLoader

Constructor Summary

Constructors

Constructor	Description
TensorFlowSavedModelEngine()

Method Summary

Modifier and Type	Method	Description
void	close()	Closes the inference engine and releases all resources.
InferenceEngine.EngineCapabilities	getCapabilities()	Gets the capabilities of this inference engine.
ModelMetadata	getMetadata()	Gets metadata about the loaded model.
InferenceResult	infer(Map<String,Object> inputs)	Performs inference on a single input.
InferenceResult	inferBatch(Map<String,Object>[] batchInputs)	Performs batch inference by running each input separately.
protected void	loadModelDirectly(ModelConfig config)	Override this method for engines that handle their own model loading.

Methods inherited from class com.codedstream.otterstream.inference.engine.LocalInferenceEngine

getModelConfig, initialize, isReady

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

TensorFlowSavedModelEngine

public TensorFlowSavedModelEngine()

Method Detail

loadModelDirectly

protected void loadModelDirectly(ModelConfig config)
                          throws InferenceException

Description copied from class: LocalInferenceEngine

Override this method for engines that handle their own model loading.

Called during LocalInferenceEngine.initialize(ModelConfig) if no ModelLoader is provided.

Overrides:

loadModelDirectly in class LocalInferenceEngine<org.tensorflow.SavedModelBundle>

Parameters:

config - model configuration containing path and options

Throws:

InferenceException - if model loading fails

infer

public InferenceResult infer(Map<String,Object> inputs)
                      throws InferenceException

Description copied from interface: InferenceEngine

Performs inference on a single input.

Specified by:

infer in interface InferenceEngine<org.tensorflow.SavedModelBundle>

Specified by:

infer in class LocalInferenceEngine<org.tensorflow.SavedModelBundle>

Parameters:

inputs - map of input name to input value

Returns:

inference result containing predictions

Throws:

InferenceException - if inference fails

inferBatch

public InferenceResult inferBatch(Map<String,Object>[] batchInputs)
                           throws InferenceException

Performs batch inference by running each input separately. Returns a list of InferenceResult objects, one per input.

Specified by:

inferBatch in interface InferenceEngine<org.tensorflow.SavedModelBundle>

Specified by:

inferBatch in class LocalInferenceEngine<org.tensorflow.SavedModelBundle>

Parameters:

batchInputs - array of input maps

Returns:

inference result containing batch predictions

Throws:

InferenceException - if inference fails

getCapabilities

public InferenceEngine.EngineCapabilities getCapabilities()

Description copied from interface: InferenceEngine

Gets the capabilities of this inference engine.

Specified by:

getCapabilities in interface InferenceEngine<org.tensorflow.SavedModelBundle>

Specified by:

getCapabilities in class LocalInferenceEngine<org.tensorflow.SavedModelBundle>

Returns:

engine capabilities (batching, GPU support, etc.)

getMetadata

public ModelMetadata getMetadata()

Description copied from interface: InferenceEngine

Gets metadata about the loaded model.

Returns:

model metadata including inputs, outputs, and format

close

public void close()
           throws InferenceException

Description copied from interface: InferenceEngine

Closes the inference engine and releases all resources.

After calling this method, the engine should not be used again.

Specified by:

close in interface InferenceEngine<org.tensorflow.SavedModelBundle>

Overrides:

close in class LocalInferenceEngine<org.tensorflow.SavedModelBundle>

Throws:

InferenceException - if cleanup fails